LiveBlog: Round Tables – Social Media & Repositories

I will be keeping occasional notes on my own Round Table session this afternoon – on Social Media and Repositories but we would welcome guest posts after the event on the other Round Tables.  Those running this afternoon are:

• What Needs to be in a package when transferring into your repository? (Chair – Ian Stuart & Theo Andrew, EDINA)
• Repositories and Linked Open Data (Chair – Adrian Stevenson, UKOLN)
• Social Media & Repositories (Chair – Nicola Osborne, EDINA)

Full notes on the Social Media session will appear here this evening. For now we are back to live blogging.

Brief reports from round tables – Facilitator – Robin Rice

Linda Kerr on the Social Media round table

We started by going round the group to see what we were interested in. SOme were tweeting deposits, some just interested

Glasgow Enlighten has a tweet button for each paper, researchers can do that to raise their profile and to comment

Mostly we talked about raising the profile and promoting items through social media. We particularly talked about Twitter and the idea that impact could be demonstrated through that sort of activity. And we also talked about not mixing up automatic tweets with public engagement type tweets and materials. Also talked about researchers and their reluctance to uptake social media and the possibility to raise the profile of materials through social media. Also talked about tapping into social networks and communities – like those on Mendeley and in other social spaces.

Usage statistics – important to get feedback. If we are using repositories what do academics find useful. Links to social networking profiles – perhaps to a researcher profile page. A way in which a researcher can raise their profile in the community – and perhaps their amazon author profiles.

And we talked about Google Scholar Citations – William retweeted a link to blog post about this – a whole new community for researchers – is that a threat or an opportunity?

Theo Andrew on SWORD packages for repository deposits.

SWORD is a proticol for depositing content into repositories. We had a very focused chat on what kind of packages we need to actually put content from point A to point B. SWORD is very simple based on AtomPUB. We should use any extensions only very sparingly. We looked at the minimum data required for data transer – really a URL would be the basic minimum. We were very concerned with how do we enoucrage repositories to share especially when repositories all do their own customisations and have differing needs. We talked about standards – can be an answer but generally more of a problem. Negotiation would be a better way to handle this. Any services for transferring content can interrogate a repository for what it understands – what metadata fields. Particularly important for our Repository Junction project which will take data and place in a series of appropriate repositories. The broker in this sense makes a lot of sense – have a relationship with a single broker.

Peter Burnhill on the Linked Open Data round table

Our topic was Linked Data and repositories. In some sense we should have asked what can linked data do for repositories, and perhaps what can repositories do for linked data. Part of this issue was about whether linked data is for the metadata or for the object. In some sense objects are all different aside from having common forms of metadata.

Motivations for linked data. To some extent it’s more about the metadata – content is often in PDF form. There was talk of giving something a URI and have everything connect to that. Then we went off into why institutional repositories should be interested in Linked Data. Partly it was about making content more accessible, another channel if you like. Another interesting idea was that this was a way of putting repositories and their content on the linked data map. But then a debate about how to make a start. How to reach a base point in using linked data. Assigning URIs or publishing minted URIs could be the way to go. Fedora definitely does this. EPrints is doing. DSpace has it in sight to do. The URIs are already there, perhaps even for the metadata they are already there. Essentially it’s about assertions without trust – a big arguement that one should just do it. And that the value far outweighed the value that might be there. And for the authors, papers etc. Although names are messy, identifiers are less messy. Identifiers for organisations easier than the any time any place people.

Q&A

Q1) What do you think about people that tweet publications or journals? Should closed materials be tweeted, should only open ones be tweeted. Surely leans to open acces smodel for greater impact.

A1) Put impact to one side, value even when non full text paper. Discovery element there. One of the things we need to consider there. Perhaps we distinguish between open access and non open access tweets. The question of Alt Metrics, tweets, etc. and impact is going to be more important. ANd the way that Google Scholar Citation works. REF will be looking at it at the narrative level of impact, not at the counts etc. Anecdotally tweeting impacts on rankings and searchability.

My comment: People do tweet about New York Times links. But there is an issue of expectation management here and we should distinguish news feeds or everything feeds from public engagement type content.

Q2 Balvier Notay, JISC) You were talking about usage statistics and I was wondering if anyone in the Social Media group mentioned the PIRUS2 project – aggregating statistics from repositories and publishers, normalising them and in harvesting statistics centrally they will do COUNTER compliance. We are looking at a statistics service type thing at a national level.